On Symmetric and Asymmetric Light Dark Matter

TL;DR

This work analyzes thermal dark matter in the 1 MeV–10 GeV range, contrasting symmetric and asymmetric scenarios under cosmological, collider, and astrophysical constraints. It shows that CMB energy-injection limits strongly restrict symmetric light DM, while ADM can evade these bounds but requires a larger annihilation cross section and a light mediator, which in turn induces self-interactions constrained by halo shapes. The authors derive a lower bound on the mediator mass from halo-ellipticity observations and map these constraints onto direct-detection cross sections for DM–nucleon and DM–electron scattering, finding that beam-dump and supernova constraints carve out much of the electron-scattering parameter space, though viable regions remain, especially for electron scattering in the light-m mediator regime. Overall, viable light ADM models require a careful balance between achieving sufficient annihilation, maintaining thermal contact with the Standard Model, and satisfying halo-shape bounds, with direct-detection prospects most promising for DM–electron interactions in the light-mediator case.

Abstract

We examine cosmological, astrophysical and collider constraints on thermal dark matter (DM) with mass mX in the range 1 MeV to 10 GeV. Cosmic microwave background (CMB) observations, which severely constrain light symmetric DM, can be evaded if the DM relic density is sufficiently asymmetric. CMB constraints require the present anti-DM to DM ratio to be less than 2*10^{-6} (10^{-1}) for DM mass mX = 1 MeV (10 GeV) with ionizing efficiency factor f ~ 1. We determine the minimum annihilation cross section for achieving these asymmetries subject to the relic density constraint; these cross sections are larger than the usual thermal annihilation cross section. On account of collider constraints, such annihilation cross sections can only be obtained by invoking light mediators. These light mediators can give rise to significant DM self-interactions, and we derive a lower bound on the mediator mass from elliptical DM halo shape constraints. We find that halo shapes require a mediator with mass mphi > 4 * 10^{-2} MeV (40 MeV) for mX = 1 MeV (10 GeV). We map all of these constraints to the parameter space of DM-electron and DM-nucleon scattering cross sections for direct detection. For DM-electron scattering, a significant fraction of the parameter space is already ruled out by beam-dump and supernova cooling constraints.

Figures (6)

• Figure 1: WMAP7 95$\%$ C.L. constraints on the DM annihilation cross section and mass for asymmetric dark matter and $s$-wave annihilation. We show constraints for various values of $r=r_\infty = \Omega_{\bar{X}}/\Omega_{X}$, the anti-DM to DM ratio at the present time. The shaded region (blue) is excluded by the WMAP7 data, with different shades corresponding to different $r_\infty$. Along the horizontal contours of constant $r$ are the values of $\langle \sigma v \rangle$ where the correct relic density can be obtained for an efficiency factor $f=1$. The turnover around $m_X \sim 10 {\rm GeV}$ comes from the drop in SM degrees of freedom when the universe has temperature $\sim1\ {\rm GeV}$. The solid red line is the intersection of the WMAP7 and relic density contours: it indicates the minimum $\langle \sigma v \rangle$ needed to obtain the observed relic density and satisfy CMB constraints for $s$-wave annihilation.
• Figure 2: (Top) Minimum $\langle \sigma v \rangle$ for efficient annihilation of the symmetric component in an ADM scenario, such that CMB bounds can be evaded, for two different values of the efficiency $f$. The black dotted line gives the thermal relic $\langle \sigma v \rangle$ for the symmetric case. (Bottom) The corresponding maximum allowed $r_\infty$, the anti-DM to DM ratio at the present time.
• Figure 3: Lower limit on the mediator mass from combining relic density and DM self-interaction constraints. We show the case of a vector mediator; the result for a scalar mediator is similar and is given in Eq. (\ref{['massbound']}). We consider DM self-interaction constraints from elliptical halo shapes and elliptical cluster shapes. Bullet cluster constraints do not give a lower bound on $m_\phi$. The dashed red line indicates the bound on the mass from elliptical halo shapes if CMB bounds are also applied, assuming efficiency $f \approx 1$.
• Figure 4: (Left) Nucleon scattering through a vector mediator. The green shaded region indicates the allowed parameter space of direct detection cross sections. The lighter green region imposes the bound of thermal coupling between the two sectors ("large width") while the larger shaded region only requires mediator decay before BBN. Also shown is the lower bound for the heavy mediator ($m_\phi \gg m_X$) case. (Right) Electron scattering through a vector mediator, for $m_\phi < m_X$ (green) and $m_\phi \gg m_X$ (red); the intersection of the two regions is shaded brown. We show the projected sensitivity of a Ge experiment, taken from Essig:2011nj. Beam dump, supernova, and halo shape constraints apply here and carve out the region of large $\sigma_e$ at low $m_X$. For more details, see the text. In the lighter green region, the condition of thermal equilibrium between the visible and hidden sectors is imposed.
• Figure 5: (Left) Constraints on mediator mass $m_\phi$ and coupling to electrons $g_e$ for $m_\phi < m_X$. The shaded region is excluded from electron anomalous magnetic moment, beam dump experiments, and supernova cooling Bjorken:2009mm. The red dashed line shows the $g_e$ value used to derive the corresponding red dashed line ("C") in the right plot. (Right) Constraints on electron scattering from Fig. \ref{['dd_SI']}. The boundaries A, B, and C are discussed in more detail in the text.